Global change contributes to drastic shifts in vegetation composition resulting in changes in ecosystem processes across the world. One important example is plant invasion, which often leads to vegetation community type conversion, such as conversion from native shrubland to invasive grassland. Chaparral, California’s most pervasive vegetation type, has recently undergone invasion. The shift from evergreen chaparral shrubs to invasive grassland will have cascading effects on ecosystem services. The overarching goal of this research is to understand water use, root and fungal dynamics of invaded chaparral communities that may inform restoration efforts. My first chapter examines how differences in root development relate to soil-water dynamics between a chaparral shrub and an invasive grass. I explored above- and below-ground strategies in concert of a native chaparral shrub and an invasive grass species in southern California using soil moisture sensors, manual minirhizotron imagery, stable isotopes, sap flux sensors and normalized difference vegetation index (NDVI). I found that the invasive grass species depleted soil moisture and produced longer roots earlier than the native shrub. Depletion of soil moisture earlier by E. calycina suggests that invasive grasses could accelerate the onset of the summer drought in chaparral systems, assuring their persistence following invasion. My second chapter examines how invasion and nitrogen deposition structure composition of fungal communities. I found that invasive grasses had a lower richness and relative abundance of symbiotic fungi compared to native shrubs. My third chapter explores how invasive-conditioned soils affect the growth of chaparral shrub seedlings and associated fungal communities and I found that native inoculation produced a more diverse fungal symbiont community. Chapter four aims to detect if invasive grass water-use strategies are detrimental to shrub seedling success, and found that invasive removal positively affected the establishment of native shrub seedlings and seedling mortality, which increased alongside invasive cover. Overall, my dissertation demonstrates that competition between invasive and native plants as well as shifts in fungal communities contributes to invasive grass persistence and shows how joining tools and perspectives from diverse fields can provide a holistic look at system responses to change.

Multiple drivers of global environmental change increasingly threaten native ecosystems, including atmospheric pollution and resulting changes in climate and nutrient cycling, and the globalization of species. These factors may also have complex and interactive ecological effects. Nitrogen (N) deposition, the input of reactive N from the atmosphere to the earth’s surface, is increasing dramatically worldwide due to anthropogenic air pollution, with the potential to negatively impact terrestrial plant diversity. Elevated N deposition may also interact with other drivers of environmental change, for example by promoting the invasion of nonnative plant species, or increasing plant susceptibility to drought or other secondary stressors. Perhaps nowhere in the U.S. is this of more immediate environmental concern than in southern California, which is a global hotspot of biodiversity and one of the most air-polluted and populous parts of the country. High levels of N deposition have been implicated in the widespread conversion of coastal sage scrub (CSS) to annual grasslands dominated by nonnative grasses and forbs. The Santa Monica Mountains National Recreation Area of southern California protects a substantial area of remaining CSS, but due to the park’s proximity to the City of Los Angeles, stands of CSS nearest urban areas may be subject to high levels of N deposition. The state of California is also in the midst of a record-breaking drought, beginning in 2011, and this may exacerbate the negative impacts of N deposition and nonnative plant species. The objective of this work is to explore the effects of N deposition, drought and nonnative plant invasion on CSS of the Santa Monica Mountains at multiple ecologically relevant scales. I explored relationships of atmospheric N pollution and N deposition with native plant richness and cover of nonnative species at the landscape level, finding N deposition reduces richness of native herbaceous species and is associated with higher nonnative cover. I also investigated the impact of multiple realistic levels of N addition on CSS in a field fertilization experiment on the low end of the N deposition gradient during a period that coincided with the California drought. Through this experiment, I demonstrated increased N availability may reduce water-use efficiency and drought tolerance of native shrubs, resulting in increased dieback, while concomitantly favoring nonnative annual species. Finally, I explored the role of the soil microbial community in mediating impacts of these factors on native and nonnative plant species, finding that N-impacted soil communities may provide less protection against drought in native shrub seedlings and increase growth of invasive plant species. Collectively, these results illustrate the significant ecological threat of increased N deposition on the severely threated CSS of southern California, and potential interactions with other drivers of global change such as extreme drought, and nonnative plant invasion.

Disturbances, natural or anthropogenic, alter ecosystem functioning by changing the biotic composition, biogeochemical cycling, or the physical structure of an environment. As carbon dioxide is a major contributor to global climate change, disturbances to different components of the carbon (C) cycle may further affect atmospheric CO2 concentrations. Soils store vast amounts of C and have the potential to sequester or release CO2. Two of the most extreme ecosystems, deserts and tropical forests, play an important role in the global C cycle, storing C in soils and vegetative biomass. The overarching objective of this dissertation is to quantify changes in soil C cycling after a disturbance in desert and tropical ecosystems. My first chapter addressed how soil inorganic C cycling changes with vegetation removal in a Colorado desert ecosystem. Through manipulative field and lab experiments, I explored changes in C pools after vegetation removal by measuring changes in the isotopic composition of C pools and effluxed CO2. I found that there were significant changes in C cycling after the vegetation was removed; there were fewer newer C inputs and that a greater proportion of CO2 comes soil inorganic C in unvegetated soils. For my second chapter, I looked at soil C recovery with revegetation of Larrea tridentata, finding that there was some recovery of the water extractable organic and microbial biomass C pools. Although the isotopic composition of soil carbon pools did not change with restoration, I also found some seasonal patterns such that changes in soil C pools may have been linked to timing of shrub and microbial activity. My third chapter addressed spatial differences in root and hyphal dynamics and production associated with soil disturbances created by leaf cutter ants (LCA). I collected continuous soil data and automated minirhizotron images to assess belowground dynamics, finding that LCA altered their soil environment and increase root and hyphal production in LCA nests. Overall, my dissertation work demonstrates that disturbances impact various components of the C cycle and that changes to C cycling processes may change how relevant those processes are at different temporal and spatial scales.

ABSTRACT OF THE DISSERTATION

Invasive Plant-Soil Feedbacks and Ecosystem Resistance and Resilience:

A Comparison of Three Vegetation Types in California

By Sara Jo Dickens

Doctor of Philosophy, Graduate Program in Plant Biology

University of California, Riverside, December 2010

Dr Edith B. Allen, Chairperson

Ecosystem processes are strongly dependant on plant-soil feedbacks. The invasion of exotic plant species can result in the introduction of novel traits capable of de-coupling native plant-soil feedbacks and leading to altered nutrient cycling and availability and microbial community composition. In general the degree to which an invading species will impact the system it invades is dependant on how much it differs from plant species native to that system. However, there are examples in which invasion of an exotic plant similar to natives has led to significant alterations of ecosystem processes. The objective of this work was to examine the impacts of a single suite of exotic annual plants invading three very different vegetation types in southern California, grasslands, coastal sage scrub and chaparral. I predicted that invasion of exotic annuals would have greater impacts on shrubland systems than grasslands due to the greater disparity in plant traits between the exotic annuals and the shrub species. Comparisons of invasion impacts, however must consider factors other than vegetation type such as soil parent material, pH and soil moisture as mechanisms by which a native system may be more or less resistant to changes associated with invasion and recover following native vegetation re-establishment. In order to examine vegetation type resistance to invasion and soil resilience of these systems, I analyzed soils for total carbon and nitrogen, extractable phosphorus and nitrogen, nitrogen mineralization, soil respiration and microbial community composition using phospholipid fatty acid analysis. Long and short-term restorations were sampled concurrently with the invaded and native vegetation types to assess resilience of soils. Regardless of vegetation type nitrate was reduced by invasion, seasonality of sampling was a greater driver of microbial community composition than invasion or restoration and abiotic factors proved to be important to microbial species composition and soil nutrient availability. Individual chemical pools and functional groups of microbes responded differently in each vegetation type. However, the degree to which invading species differed from natives did not predict the level to which invasion would impact the system.

Understanding the composition of ecological communities that arise from potential species pools has implications for community assembly and applications for restoration. Invasive species pose special challenges to restoration by contributing to ecosystem degradation as well as resisting restoration efforts. In the face of such challenges, understanding the complex of mechanisms working together to enable an invasive species to establish and spread may lead to better management strategies and greater restoration success. The overall objective of this dissertation is to understand mechanisms contributing to the success of a Mediterranean annual grass, Bromus diandrus, through the use of both field and greenhouse studies, and to use this understanding to inform restoration of invaded ecosystems. More specifically, I consider three potential mechanisms of invasion: 1) plant functional traits, 2) plant-soil feedback, and 3) soil legacy effects. The results of the studies of this dissertation demonstrate that multiple mechanisms of invasion promote Bromus diandrus success. First, functionally similar native plant communities did not demonstrate biotic resistance to B. diandrus invasion during restoration studies. Rather, earlier germination and larger seed mass of B. diandrus allows this invasive grass to establish even in the presence of morphologically similar native species with greater relative growth rates. Second, positive plant-soil feedback in B. diandrus attributed to the fine arbuscular mycorrhizal fungi contributes to its overall success. Lastly, strong soil legacies in abandoned agriculture also contribute to B. diandrus invasion and inhibit successful reestablishment of native plants. Root fungal pathogens found in abandoned agricultural fields result in decreased biomass of some native species as well as B. diandrus. A greater understanding of the mechanisms contributing to B diandrus invasion success suggests that restoration attempts should seed with functionally similar natives while manipulating germination cues and utilize facilitated microbial inoculations to reduce Bromus diandrus establishment. However, many mechanisms contribute to the overall success of this invasive species making it competitively superior, and eradication of B. diandrus on a large scale is unlikely.

Undisturbed wildland ecosystems are impacted by anthropogenic nitrogen (N) emissions being deposited significant distances from their sources. The main sources of atmospheric N inputs in the United States include industrial and automotive exhaust and emissions from agricultural wastes and fertilizers. Emission sources can be differentiated by analyzing the δ18O and δ15N of emitted compounds. Through a combination of field and laboratory studies, this dissertation aims to identify the various emission sources impacting the western Sonoran Desert and to determine how they are altering plant available nitrogen in the region. The first objective of this research was to measure if fractionation of HNO3 δ18O and δ15N occurs to filters of commonly used ambient HNO3 collectors. These collectors were then placed along a N deposition gradient to measure the isoscapes of HNO3 δ15N and δ18O and extrapolate how each emission source contributes to regional anthropogenic N. Lastly, soil and plant tissue were collected at each site to evaluate whether atmospheric patterns were conserved through the ecosystem.

Results suggest that ambient HNO3 does not fractionate isotopically when binding to the passive sampler filters when exposed in a continuous stirred tank reactor. The HNO3 δ18O and δ15N were within 0.5 / of the source HNO3 when exposed to controlled high (20 μg m^-3) and low (10 μg m^-3) concentrations for four weeks. When exposed under field conditions, the samplers verified that anthropogenic nitrogen impacting the Coachella Valley came from two sources based on the changes to HNO3 δ18O and δ15N across the area; vehicle emissions from the Los Angeles air basin and agricultural emissions from around the Salton Sea. There was also a distinct separation in values for sites within Joshua Tree National Park suggesting that the Little San Bernardino Mountains act as effective barrier from air pollution moving in the park. Finally, surface soil NO3- was the most effective indicator of anthropogenic additions; with NO3- concentrations and δ18O linearly tracking atmospheric HNO3 concentrations. The leaf tissue δ15N of the regionally dominant shrub, Larrea tridentata, also decreased at sites with increasing anthropogenic inputs. Both the atmospheric samplers and soil surface N analyses will provide land managers with effective tools to quickly identify regions of high anthropogenic inputs in a desert environment.